This paper formulates one-dimensional (1-D), recursive, multiplicative time series models for digital images and demonstrates their use for adaptive predictive coding of such images. The performance of the scheme pres...
详细信息
This paper formulates one-dimensional (1-D), recursive, multiplicative time series models for digital images and demonstrates their use for adaptive predictive coding of such images. The performance of the scheme presented here is superior compared to that of conventional 1-D modelling techniques, because correlation among all neighboring pixels of interest can be taken into account. Further, the projection-based constrained least squares identification technique proposed here guarantees stability of the underlying predictor, which makes the scheme more robust compared to the ones that use 2-D recursive models where predictor stability cannot be guaranteed.
In this paper a new fuzzy logic-based lossy predictive coding system for gray-scale still image compression is developed. The proposed coder employs a recently introduced adaptive fuzzy prediction methodology in the p...
详细信息
In this paper a new fuzzy logic-based lossy predictive coding system for gray-scale still image compression is developed. The proposed coder employs a recently introduced adaptive fuzzy prediction methodology in the predictor design. In addition, it adopts a novel fuzzy gradient-adaptive quantization scheme. The proposed coding technique possesses superior performance over its non-fuzzy linear counterparts especially at low bit quantization. This is due to the inherent adaptivity in the fuzzy prediction methodology as well as the gradient-adaptive quantization scheme. Simulation results are provided to demonstrate the efficient performance of the proposed fuzzy predictive coding system.
We present a predictive neural network called neural predictive coding (NPC). This model is used for nonlinear discriminant features extraction applied to phoneme recognition. We validate the nonlinear prediction impr...
详细信息
We present a predictive neural network called neural predictive coding (NPC). This model is used for nonlinear discriminant features extraction applied to phoneme recognition. We validate the nonlinear prediction improvement of the NPC model. We also, present a new extension of the NPC model: NPC-3. In order to evaluate the performances of the NPC-3 model, we carried out a study of Darpa-Timit phonemes (in particular /b/, /d/, /g/ and /p/, /t/, /q/ phonemes) recognition. Comparisons with traditional coding methods are presented. We also show how an adaptative constraint allows improvements on the recognition task.
The adaptive predictive coding with transform domain quantization (APC-TQ) technique was proposed by Bhaskar (1991) for the compression of audio signals. Since then, significant developments have taken place leading t...
详细信息
The adaptive predictive coding with transform domain quantization (APC-TQ) technique was proposed by Bhaskar (1991) for the compression of audio signals. Since then, significant developments have taken place leading to a reduction in the coding rate. While enhancing the audio quality. These developments include (i) the use of block size adaptation to exploit the variations in the stationarity of the signal, (ii) high resolution spectral modeling using LPC analysis orders up to 64, and (iii) an adaptive bit-allocation procedure to minimize coding noise power as well as minimize the perception of coding noise. The result is a near transparent quality compression of 5 kHz bandwidth audio at a rate of 17 kbit/s. This technology will find applications in the distribution and transmission of AM quality audio programming over low rate channels such as the INMARSAT Standard A, B and aeronautical systems.< >
predictive coding, adapted from text categorization for litigation support, is an evolving process with identification of responsive documents and changing labeling decisions. The current state-of-art within predictiv...
详细信息
predictive coding, adapted from text categorization for litigation support, is an evolving process with identification of responsive documents and changing labeling decisions. The current state-of-art within predictive coding workflow uses Active Learning, where a new model is periodically rebuilt with additional documents reviewed, to continuously revise a model and improve the identification of responsive documents. We propose an alternative approach to recursively update the model using the Unscented Kalman Filter for each additional labeled document. With synthetic text streaming data and induced concept drift, we show that our approach learns new patterns at a faster rate, renders better accuracy and recall, and requires a reduced labeling cost, which when combined makes it potentially a better alternative in updating the model in the setting of Active Learning for predictive coding.
In recent years, transformer-based, large-scale language models have greatly advanced deep learning in NLP tasks. These models allow transfer learning to be performed on NLP in similar methods to what was previously d...
详细信息
ISBN:
(纸本)9781665480468
In recent years, transformer-based, large-scale language models have greatly advanced deep learning in NLP tasks. These models allow transfer learning to be performed on NLP in similar methods to what was previously done for computer vision. Two of the latest models are DistilBERT and Longformer -- the former is a distilled version of BERT, which enables faster training and inferencing on off-cloud servers, and the latter offers capability of working with long texts. In this paper, we study empirical comparisons of the effectiveness of the two deep learning methods along with a logistical regression method for text classification, using three real-world datasets from legal document reviews. The study shows that Longformer performs better (up to 10%) than or at par with the other two methods. Cross-dataset evaluation is leveraged as well to validate the performance of Longformer as a viable method when labeled data is not available.
Digital pathology tasks have benefited greatly from modern deep learning algorithms. However, their need for large quantities of annotated data has been identified as a key challenge. This need for data can be counter...
详细信息
Digital pathology tasks have benefited greatly from modern deep learning algorithms. However, their need for large quantities of annotated data has been identified as a key challenge. This need for data can be countered by using unsupervised learning in situations where data are abundant but access to annotations is limited. Feature representations learned from un-annotated data using contrastive predictive coding (CPC) have been shown to enable classifiers to obtain state of the art performance from relatively small amounts of annotated computer vision data. We present a modification to the CPC framework for use with digital pathology patches. This is achieved by introducing an alternative mask for building the latent context and using a multi-directional PixelCNN autoregressor. To demonstrate our proposed method we learn feature representations from the Patch Camelyon histology dataset. We show that our proposed modification can yield improved deep classification of histology patches.
Compressive sensing (CS) is beneficial for unmanned reconnaissance systems to obtain high-quality images with limited resources. The existing prediction methods for block-based compressive sensing of images can be reg...
详细信息
ISBN:
(数字)9781728180250
ISBN:
(纸本)9781728180267
Compressive sensing (CS) is beneficial for unmanned reconnaissance systems to obtain high-quality images with limited resources. The existing prediction methods for block-based compressive sensing of images can be regarded as the particular coefficients of weighted predictive coding. To find better prediction coefficients for BCS, this paper proposes two weighted prediction methods. The first method converts the prediction model of measurements into a prediction model of image blocks. The prediction weights are obtained by training the prediction model of image blocks offline, which avoiding the influence of the sampling rates on the prediction model of measurements. Another method is to calculate the prediction coefficients adaptively based on the average energy of measurements, which can adjust the weights based on the measurements. Compared with existing methods, the proposed prediction methods for BCS of images can further improve the reconstruction image quality.
predictive coding, the term used in the legal industry for document classification using machine learning, presents additional challenges when the dataset comprises instant messages, due to their informal nature and s...
详细信息
predictive coding, the term used in the legal industry for document classification using machine learning, presents additional challenges when the dataset comprises instant messages, due to their informal nature and smaller sizes. In this paper, we exploit a data management workflow to group messages into day chats, followed by feature selection and a logistic regression classifier to provide an economically feasible predictive coding solution. We also improve the solution's baseline model performance by dimensionality reduction, with focus on quantitative features. We test our methodology on an Instant Bloomberg dataset, rich in quantitative information. In parallel, we provide an example of the cost savings of our approach.
The improvement of multimedia, biomedical, medical and digital imaging has led to huge amount of data required to represent modern imagery. This requires a large disk space for storage and more time for transmitting f...
详细信息
The improvement of multimedia, biomedical, medical and digital imaging has led to huge amount of data required to represent modern imagery. This requires a large disk space for storage and more time for transmitting from one computer networks to the other. Both are relatively expensive. In the biomedical images, the resolution is also very high. So, the factors like large space for storage, time and high resolution proves the need for images compression. Image compression addresses those problems and gives the solution of reducing the space required to represent biomedical image and then this takes less time to transmission from one computer network to other and it converts high resolution into the low resolution. Biomedical images are in the form of DICOM. In this paper we consider the condition of medical report. If we send the biomedical image of person from one place to other by using the network, then we will perform the compression techniques. In this paper we also perform the inverse transform for encoding or decoding the image. In this paper we propose the predictive compression algorithm with the help of ROI on biomedical images.
暂无评论